The djozz-lumen is the output of my Sunwayman D40A on high setting divided by 550. It was established a few years ago and I keep it as my standard because it is consistent over all my measurements, and with the D40A always available it is easy to maintain.
Since then I have learned that the djozz-lumen is likely higher than the real lumen, some comparisons that I did in Germany with flashlights measured before in the LedLenser sphere, and an exchange of 2 different flashlights with maukka suggest that the djozz-lumen is on average around 9% high. But compared to maukka’s setup a high CRI 4000K flashlight (the 348 light that maukka is selling now for calibration) produced a bigger difference (11) than a 5000K 70CRI Nitecore P12 (7), so despite using a very high quality well V-lambda corrected luxmeter I have a spectrum sensitivity difference with maukka as well.
And lots of people think it’s easy to measure lumens. Just buy a lumen meter on ebay, they say. :person_facepalming: It’s funny.
I think the general consensus is that warmer tints do read a bit lower in lux compared to cooler tints, but this is still debated. I try not to think too much about it. It makes my head hurt. Lol
I’ve considered buying a calibration source from maukka. Before doing that though, I’d need to build a more proper integrating sphere… and I’m not sure I care enough to do that.
So I have a very cheap and compact measurement device with calibration based on a couple Zebralight SC52s, selfbuilt’s reviews, and a few people who were publishing numbers on BLF a few years ago. And it almost certainly reads between 10% and 50% higher than an ANSI lumen.
The downside is that it’s known to be skewed. The upside is that it typically matches most manufacturers’ published specs pretty closely. But then, another downside is that, apparently, most manufacturers exaggerate.
Also, since what I’m using is small, it can’t handle more than about 2500 lm without some sort of attenuator. So I get to mess with neutral density filters while measuring bright lights.
But back on topic, it’ll be interesting to see how the GT-Mini performs. It seems like it’ll be a D1S with a nicer host, flatter but slightly lower output, and a more full-featured UI.
Have you considered getting a TA Lumen Tube? When you add up the time and parts cost to build a device that properly diffuses both floody and throwy lights plus an additional $75 for the calibration lights, you actually come out ahead just buying his for about $150 shipped. It’s well thought out and very easy to use. You don’t even need a calculator.
Do we have an estimate on when the first units are supposed to start shipping? I kinda want to see beamshots of both tints before placing my order. I would have preferred a 5000k option, so I’m not sure which one will be closer to my preferences.
Just filled out the form, is the GB still on Martin?
About the driver: is it sure a buck driver, as mentioned in this thread? I can’t believe they went for that, the voltage difference between a 18650 cell and a XP-L Hi is so small that it will only regulate current a little in the first few minutes, I’d choose for simple cheap lineair and have that extra boost of direct drive in those minutes. I’m curious about the first pictures of the driver.
Looking at Texas Ace’s measurements, the XP-L HI has a V_f of around 3.7V @ 4.5A.
So accounting for the voltage sag under load, a high drain cell should stay in the regulated region for slightly less than half of its capacity on turbo.
The other big advantage of a regulated buck driver over a FET is improved efficiency in the intermediate modes. So it should be able to produce noticeably better runtimes at 60-80% of max than the Emisar D1S. I’ll gladly take that over an extra 100 turbo lumens that won’t be noticeable by the eye anyway.
Regulated about half of its capacity, is this correct? A 30Q at 5A starts between 3.9 and 4 V, at 8 minutes it drops below 3.8V, at half capacity it is at 3.6V. Between cell and led there is always a voltage drop caused by body, springs and wiring, if you get a 0.2V drop (at 5A current), that is very good, in that case the springs are already bypassed. So the battery voltage comes very soon very close to the led voltage for 5A. Further from what I understand about buck drivers, they need a minimum overhead voltage, a value that I can find is a minimum 0.5V difference needed to function. So I do not see a buck driver doing a thing here.
But I must confess that my knowledge about basic calculations is better than my knowledge about drivers…
Wow. 40. If that is true then I am going to have to rethink a lot about manufactures numbers. I was under the impression (perhaps falsely) that the zebralight calibration was one of the more premuim methods. It was for that reason that I assumed Acebeam L16, X45, X80, all of my olights and many other top brands were within about a 10 margin of error from the truth. Most often with those brands I would get just under or just over the rated output.
I have ordered a set of Maukkas lights and could hardly be more excited!
.
.
.
.
About the GT MINI. I get why 18650 was the choice. That said if 20700/21700 options come available I would be crazy over that.
